Vol. 169, No. 12JOURNAL OF BACTERIOLOGY, Dec. 1987, p. 5653-56620021-9193/87/125653-10$02.00/0Copyright © 1987, American Society for Microbiology

Glyceraldehyde-3-Phosphate Dehydrogenase Gene from Zymomonasmobilis: Cloning, Sequencing, and Identification of Promoter Region

T. CONWAY, G. W. SEWELL, AND L. 0. INGRAM*Department of Microbiology and Cell Science, University of Florida, Gainesville, Florida 32611

Received 28 May 1987/Accepted 18 September 1987

The gene encoding glyceraldehyde-3-phosphate dehydrogenase was isolated from a library of Zymomonasmobilis DNA fragments by complementing a deficient strain of Escherichia coli. It contained tandem promoterswhich were recognized by E. coli but appeared to function less efficiently than the enteric lac promoter in E.coli. The open reading frame for this gene encoded 337 amino acids with an aggregate molecular weight of36,099 (including the N-terminal methionine). The primary amino acid sequence for this gene had considerablefunctional homology and amino acid identity with other eucaryotic and bacterial genes. Based on thiscomparison, the gap gene from Z. mobilis appeared to be most closely related to that of the thermophilicbacteria and to the chloroplast isozymes. Comparison of this gene with other glycolytic enzymes from Z. mobilisrevealed a conserved pattern of codon bias and several common features of gene structure. A tentativetranscriptional consensus sequence is proposed for Z. mobilis based on comparison of the five known promotersfor three glycolytic enzymes.

Zymomonas mobilis is an obligately fermentative, gram-negative bacterium which cannot grow in the absence of afermentable carbohydrate (24, 35). In nature, this organismis found in honey and plant saps and is a common contam-inant of commercial ethanolic fermentations. All isolates ofthis genus exhibit the ability to metabolize only a verylimited range of carbohydrates, including glucose, fructose,and sucrose. These sugars are metabolized via the enzymesof the Entner-Doudoroff pathway in Z. mobilis, with theproduction of ethanol and carbon dioxide as waste products.The entire energy needs of this organism (homeostasis,biosynthesis, and growth) are dependent on the net produc-tion of a single ATP per molecule of hexose consumed (24).Despite an apparent inefficiency of ATP production, thisorganism is capable of rapid growth, with generation times of90 to 120 min (28), and exhibits a higher rate of glycolyticflux and ethanol production than Saccharomyces cerevisiae(13, 28).The enzymes of glycolysis are particularly abundant in Z.

mobilis and are estimated to represent over half of thesoluble protein (1). In this organism, substrate-level phos-phorylations by pyruvate kinase and glyceraldehyde-3-phosphate dehydrogenase occupy particularly central roles.Previous studies in our laboratory have compared the spe-cific activities of these and other glycolytic enzymes in Z.mobilis with the rates of glycolytic flux at different stages ofgrowth and fermentation during batch culture (27). In gen-eral, the activities of all glycolytic enzymes appeared to be inexcess from in vitro assays under substrate-saturating con-ditions. The specific activities of all enzymes except pyru-vate kinase and glyceraldehyde-3-phosphate dehydrogenasecontinued to increase after the exponential phase as the ratesof glycolytic flux declined. In contrast, the changes in thespecific activities of these two enzymes were directly corre-lated with changes in the rates of glycolytic flux.

* Corresponding author.

Glyceraldehyde-3-phosphate dehydrogenase catalyzes thereversible oxidation and phosphorylation of glyceraldehyde-3-phosphate to produce 1,3-diphosphoglycerate (15). Thisenzyme (in association with 1,3-diphosphoglycerate kinase)represents a common link between ATP production and theextent of NAD+ reduction and thus occupies a uniqueposition in glycolysis. In Z. mobilis (29) and other organisms(15), this enzyme is composed of four identical subunits. Ithas been purified to homogeneity from Z. mobilis and isreported to represent over 5% of the protein in solubleextracts (29). Similarly high levels of this enzyme have beenreported for mammalian muscle tissue and fermenting S.cerevisiae (15). The monomer molecular weight of the Z.mobilis enzyme has been estimated to be between 32,500 and41,000 (29); those from other organisms are typically 36,000(15). The kinetics of glyceraldehyde-3-phosphate dehydro-genases are complex and typically do not follow simpleMichaelis-Menten kinetics. This enzyme represents an im-portant model system for investigations of protein evolution,structure, and function (15, 33, 36). The primary structures

TABLE 1. Plasmids and strains used

Strain or plasmid Relevant genotype Source or reference

Z. mobilis CP4 Prototroph 28

E. coliDF221 gap 16TC4 recA 7DH5ot AlacZM15 recA BRLaJM10o AIacZM15 F' BRL

PlasmidspLOI193 cat tet 7pLOI310 gap cat This studypLOI311 gap bla This studypLOI312 gap bla This studypLOI314 gap bla This studypUC18 bla lacI'Zb BRL

a BRL, Bethesda Research Laboratories (Gaithersburg, Md.).bIncomplete lacl and incomplete lacZ.

5653

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

5654 CONWAY ET AL.

ApL01310Bam.SIj B

I v Iv I e R PvI g

Pv Saui'BunI

-d I I a IaA . oh

O i

pL01311PI

4 -.I NU!-r ac

lrIupLOI 314

pL01312RFv P

I

5 '5.5

7V2.8

P FHle?

I iI * iiI|

1.7

plac

FIG. 1. Restriction maps ofDNA fragments (A) and plasmids (B) containing the gap gene from Z. mobilis (Zm). The bold arrow representsthe gap gene. No BamHI sites were retained at the junctions of insert and vector DNA in the original construction (pLOI310). The gap genewas localized on a 2.8-kb EcoRI fragment within pLOI310, and this fragment cloned into the EcoRI site of pUC18 to produce pLOI311.Deletions of this 2.8-kb fragment were subloned into the SmaI site of pUC18 in both orientations with respect to the lac promoter (pLOI314and pLOI312). Abbreviations: Cmr, phloramphenicol resistance gene; Apr, ampicillin resistance gene; ori, origin of replication from eitherRSF1010 (oriV) or ColEl; Bam, BamHI; Sau, Sau3A; B, BstEII; R, EcoRI; P, PstI; H, HindlIl; Pv, PvuII; F, site of blunt fusion. Numbersunder lines denote the DNA size (in kilobases).

and DNA sequences of this enzyme have been determined inmany organisms, and all have considerable amino acidhomology. Three-dimensional structures of several of thesehave been established and clearly identify conserved cleftsinvolved in substrate binding, cofactor binding, and subunitinteraction (6, 15, 36).

In this study we have cloned and sequenced glyceral-dehyde-3-phosphate dehydrogenase (EC 1.2.1.12) from Z.mobilis as a first step in our investigation of the control of itsexpression.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions. Thestrains and plasmids used in this study are listed in Table 1.Z. mobilis CP4 was grown at 30°C in complex medium (28)containing 20 g of glucose per liter. Escherichia coli strainsTC4 and DH5a were grown in Luria broth (21) or on Luriaagar (1.5% agar) lacking added carbohydrate. Strain DF221(CGSC 5584) was grown in minimal medium as described byHillman and Fraenkel (16), containing 4 g of sodium suc-

-A ME 0 m 0 m

1. I 1111=M -1 I --- L-- W--"% I "I -1- I

J. BACTERIOL.

vtp I

IW ',¢

.3

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

Z. MOBILIS GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE

TABLE 2. Specific activity of glyceraldehyde-3-phosphatedehydrogenase clones

Glyceraldehyde-3-phosphateE. coli strain Plasmid dehydrogenase sp act

(IU/mg)

DF221 None <0.02TC4 None 0.55DF221 pLOI310 0.46

pLOI312 0.25pLOI314 4.53

Enzymes and chemicals. Restriction enzymes were ob-tained from Bethesda Research Laboratories (Gaithersburg,Md.) and from International Biotechnologies, Inc. Reversetranscriptase from avian myeloblastosis virus and othernucleic acid-modifying enzymes were purchased fromBethesda Research Laboratories. Radioactive compoundswere purchased from New England Nuclear Corp. (Boston,Mass.). Glyceraldehyde-3-phosphate was purchased fromSigma Chemical Co. (St. Louis, Mo.).

RESULTS AND DISCUSSION

cinate and 1 g of glycerol per liter. Carbohydrates weresterilized by filtration and added to autoclaved basal mediumafter cooling. Selections for complementation of theglyceraldehyde-3-phosphate dehydrogenase defect in strainDF221 were made on Luria agar lacking added carbohydrateand on minimal plates containing glucose (16) with appropri-ate antibiotics. Ampicillin (50 mg/liter) and chloramphenicol(40 mg/liter) were used to select recombinants of E. coli;5-bromo-4-chloro-3-indolyl-p-D-galactopyranoside (20 mg/li-ter) was used to identify DNA insertions which inactivatedP-galactosidase activity in strain DHSt containing deriva-tives of pUC18. Strains of E. coli were transformed by themethod of Mandel and Higa (22).

Cloning the gene encoding glyceraldehyde-3-phosphatedehydrogenase. A library of Z. mobilis chromosomal DNAwas constructed previously (10) by inserting 5- to 7-kilobase(kb) fragments from a Sau3A partial digest into the BamHIsite of the tetracycline resistance gene in pLOI193 (7). Theresulting plasmids retained chloramphenicol resistance as aselectable marker. Plasmids were prepared from the pooledcolonies of the original transformation by the alkaline lysisprocedure (23). These plasmid preparations were used as asource of Z. mobilis DNA for complementation of theglyceraldehyde-3-phosphate dehydrogenase defect in E. coliDF221. Strain DF221 does not grow on Luria agar lackingcarbohydrate or on glucose minimal medium (16).

Assay of glyceraldehyde-3-phosphate dehydrogenase. Cellswere harvested and disrupted as described previously (10).Enzyme activity was determined by the arsenolysis proce-dure (5). Activities are expressed as international units permilligram of total cell protein. Protein was determined by themethod of Lowry et al. (20) as described by Layne (19), withbovine serum albumin as a standard.Sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

Gels were prepared and stained as described previously (9).Determination of DNA sequence. The gap gene from Z.

mobilis was sequenced by a modification (8) of the dideoxymethod of Sanger et al. (32). The entire sequence of theinsert contained in pLOI312 was sequenced in both direc-tions by using an overlapping set of Bal3l-generated dele-tions which had been cloned into derivatives of phage M13.Sequence data were analyzed with the programs of Pustelland Kafatos (30), obtained from International Biotechnolo-gies, Inc. (New Haven, Conn.).

Analysis of transcriptional initiation. The 5' termini oftranscripts were mapped by the primer extension method asdescribed previously (9). The primer used for these experi-ments was synthesized with an Autogen 500 oligonucleotidesynthesizer (Millipore/Genetic Design, Bedford, Mass.) andwas complementary to the noncoding strand of the gap gene(CAACGCTAATTACCG) spanning (3') base pair (bp) 169through (5') bp 183. A sequencing ladder was prepared withthe same primer to map the transcriptional starts.

Cloning of the gene encoding glyceraldehyde-3-phosphatedehydrogenase (gap). Plasmids containing a library of Z.mobilis CP4 DNA were transformed into E. coli DF221.Transformants were selected for growth on Luria brothplates containing chloramphenicol. Ten clones were isolatedfrom plates which contained approximately 10,000 chloram-phenicol-resistant recombinants (estimated from sampleplating on minimal medium containing chloramphenicol,glycerol, and succinate). Based on digestions with PvuII,these fell into three groups and contained 4- to 6-kb inserts.One of these, designated pLOI310, was chosen for furtherstudy (Fig. 1).

Plasmids were immediately transferred into strain TC4 orstrain DH5a for further study to avoid potential rearrange-ments and concatenation, which occur in strain DF221.Subclones were constructed in the polylinker region ofpUC18 and transformed into strain DH5ao by using theinsertional inactivation of P-galactosidase activity as amarker. Plasmids were isolated and checked for the pres-ence of a functional Z. mobilis gap gene by transformationinto strain DF221. Both HindlIl and BstEII were initiallyidentified as being within the Z. mobilis gap gene duringsubcloning. This gene was subsequently localized within a2.8-kb EcoRI fragment (pLOI311). This EcoRI fragment wasisolated from an agarose gel, treated for increasing timeswith Bal3l (23), and ligated into the SmaI site of pUC18.Representative clones containing inserts of decreasinglength were tested for their ability to complement the gapdefect in strain DF221. Two clones were chosen for furtherstudy, pLOI312 and pLOI314. The inserts in plasmidspLOI312 (1.7 kb) and pLOI314 (2.0 kb) are in oppositeorientation with respect to the lac promoter (Fig. 1).

Expression of glyceraldehyde-3-phosphate dehydrogenase

-66

*p_ --45

-̂~ -35

-24

a b c d e fFIG. 2. Sodium dodecyl sulfate-polyacrylamide gel stained with

Coomassie blue showing expression of the Z. mobilis gap gene in E.coli. Lanes: a, strain DF221; b, strain TC4(pUC18); c, strain TC4(pLOI314); d, strain TC4(pLOI310); e, strain TC4(pLOI312); f, Z.mobilis CP4. The positions of molecular weight markers (in thou-sands) are shown to the right.

VOL. 169, 1987 5655

-, f-I-A"" "'

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

5656 CONWAY ET AL. J. BACTERIOL.

-266 -156* * * * * * * * * * *GTTTCCTTGGCTGTTGAGGTCGCAGACTCTAGAAACAGCAGGTTTTGGGGCAGATGTCGTGTcGATGccGAGTTGGAcTTTGTTcGATcAACAACCCGAAT*

-166 -50* * * * * * * * * *

CCTATCGTAATGATGTTTTGCCCGATCAGCCTCAATCGACAATTTTACGCGTTTCGATCGAAGCAGGGACGACAATTGGCTGGGAACGGTATACTGGAATAAAn>>>-

1 50* * * * * * * * * *

TGGTCTTCGTTATGG AOg&TGTTTGGTGCATCGGCCCCGGCGAATGATCTATATGCTCATTTCGGCTTGACCGCAGTCGGCATCACGAACAAGGTGTTG

* * * * * * * 8.D.* *GCCGCGATCGCCGGTAAGTCGGCACGTTAAAAAATAGCTATGGAATATAATAGCTACTTAATAAGTTAGGAGAATAAAC ATG GCG GTT AAA GTT GCG

fMet Ala Val Lys Val Ala

200 250a * * a * * * *

ATT AAT GGC TTC GGC CGT ATC GGC CGT CTG GCT GCG CGT GCA ATT CTG TCG CGT CCG GAT TCT GGA CTT GAA CTG GTTII* Asn Gly Pho Gly Arg I1e Gly Arg Lou Ala Ala Arg Ala I1e Lou Ser Arg Pro Asp Ser Gly Lou Glu Leu Val

300

ACC ATC AAT GAC CTG GGT AGC GTT GAA GGC AAC GCT TTC TTG TTC AAG CGT GAC AGT GCA CAT GGC ACC TAT CCG GGTThr Ile Asn Asp Lou Gly Ser Val Glu Gly Asn Ala Pho Lou Phe Lys Arg Asp Ser Ala His Gly Thr Tyr Pro Gly

356 400

ACC GTT ACC ACC GAA GGT AAC GAC ATG GTC ATC GAT GGC AAG AAA ATT GTT GTC ACC GCT GAA CGT GAT CCG GCC AATThr Val Thr Thr Glu Gly Asn Asp Met Val Ile Asp Gly Lys Lys Ile Val Val Thr Ala Glu Arg Asp Pro Ala Asn

450

CTG CCG CAC AAA AAG CTT GGC GTT GAT ATC GTG ATG GAA TGC ACG GGT ATC TTC ACC AAC ACC GAA AAA GCT TCT GCTLou Pro His Lys Lys Lou Gly Val Asp I1e Val Met Glu Cys Thr Gly 110 Phe Thr Asn Thr Glu Lys Ala Ser Ala

500 550

CAT CTG ACC GCA GGT GCC AAG AAG GTT CTG ATT TCT GCT CCG GCT AAA GGC GAT GTT GAT CGC ACC GTT GTT TAT GGTHis Leu Thr Ala Gly Ala Lys Lys Val Lou Ile Ser Ala Pro Ala Lys Gly Asp Val Asp Arg Thr Val Val Tyr Gly

606

GTG AAC CAC AAA GAC CTG ACC GCA GAT GAC AAG ATC GTT TCC AAC GCA TCT TGC ACC ACC AAC TGC TTG GCT CCG GTTVal Asn His Lys Asp Lou Thr Ala Asp Asp Lys I1e Val Ser Asn Ala Ser Cys Thr Thr Asn Cys Leu Ala Pro Val

656 700

CTG CAT GTT CTT CAG CAG AAA ATC GGT ATC GTT CGT GGT TTG ATG ACC ACC GTT CAC AGC TTC ACC AAC GAT CAG CGCLou His Val Lou Gln Gln Lys I1e Gly Ile Val Arg Gly Lou Met Thr Thr Val His Ser Phe Thr Asn Asp Gln Ary

750

ATC CTC GAT CAG ATC CAT TCT GAT CTG CGT CGT GCC CGC ACG GCC AGC GCT TCC ATG ATC CCG ACC AGC ACC GGT GCTIl Lou Asp Gln I10 His Ser Asp Lou Arg Arg Ala Arg Thr Ala Ser Ala Ser Met I1e Pro Thr Ser Thr Gly Ala866 856GCC CGC GCT GTT GCT CTC GTG ATC CCG GAA CTC AAA GGT AAG CTG GAT GGT ATC TCC ATC CGC GTC CCG ACC CCG GATAla Arg Ala Val Ala Lou Val Ile Pro Glu Lou Lys Gly Lys Lou Asp Gly Ile Ser I1e Arg Val Pro Thr Pro Asp

900 956

GTC AGC CTT GTT GAC TTC ACC TTC GTT CCG CAG CGT GAC ACG ACC GCA GAA GAA ATC AAC AGC GTT CTT AAA GCC GCTVal SBr Lou Val Asp Phe Thr Pho Val Pro Gln Arg Asp Thr Thr Ala Glu Glu Ile Asn Ser Val Lou Lys Ala Ala

1666GCT GAC ACC GGT GAT ATG ACT GGC GTT CTC GGT TAC ACC GAC GAA CCC CTC UTT TCC CGT GAC TTC TAC TCC GAT CCGAla Asp Thr Gly Asp Mot Thr Gly Val Lou Gly Tyr Thr Asp Glu Pro Lou Val Ser Arg Asp Phe Tyr Ser Asp Pro

1050 1166

CAC AGC TCC ACC GTT GAC AGC CGT GAA ACG GCC GTT CTC GAA GGT AAG CTG GCT CGT GTT GTT GCA TGG TAT GAC AACHis Ber Ser Thr Val Asp Ser Arg Glu Thr Ala Val Lou Glu Gly Lys Lou Ala Arg Vol Val Ala Trp Tyr Asp Asn

1150

GAA TGG GGC TTC TCC AAT CGT ATG GTC GAT ACC GCT GCT CAG ATG GCT AAA ACG CTC TAO TTTTTGCGATCTTATCGGTAAAAGlu Trp Gly Phe Sor Asn Arg Mat Val Asp Thr Ala Ala Gln Met Ala Lys Thr Lou

1200 1250GGGTTGTCCCGCTTTTTGTCCTTGTGGCAAGGGGCGGGATGCCTCTTAATTACTGCAGTTAGTTATTAAGCTGATTGCAGCATTAACCTCCTGATACAGGCCC

1300 1350

GTTATGGGATATATCCGTTATAGCTGGGTCTCGTAAGGAGGCTGTCTCCGTTATCAGATCCGCTTTATrGATATAAGGCGGCCAAAAGGAGGATATAATGGCTT1400 1450

TTCGTACATTAGATGATATTGGTGACGTCAAAGGTAAGCGCGTTCTTGTTCGCGAAGATCTGAACGTTCCTATGGATGGCGATC

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

Z. MOBILIS GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE

TABLE 3. Comparison of codon usage

Amino E. coli Z. mobilisbacid Codona combinedb.c

(mol%) gap pdc adhB Combinedd

Phe TTTTTC

Leu TTATTGCTTCTCCTACTG

Ile ATTATCATA

Met ATG

Val GTTGTCGTAGTG

Ser TCTTCCTCATCGAGTAGC

Pro CCTCCC

CCACCG

Thr ACTACCACAACG

Ala GCTGCCGCA

GCGTyr TAT

TAC

His CATCAC

Gln CAACAG

Asn AATAAC

Lys AAAAAT

Asp GATGAC

Glu GAAGAG

1.3 0 0.4 0.8 0.42.2 2.7 2.9 2.1 2.6

0.7 0 0.4 0 0.10.9 0.9 1.8 1.3 1.30.8 1.5 1.6 1.8 1.60.8 2.1 1.8 1.3 1.70.2 0 0 0 06.8 3.6 3.4 5.2 4.1

2.2 1.2 1.1 1.6 1.33.7 4.5 3.6 3.1 3.70.2 0 0 0 0

2.8 2.4 2.3 4.4 3.0

2.9 7.71.2 1.51.8 02.2 0.9

4.1 5.03.2 2.30 0

0.5 0.5

5.62.30

0.6

1.3 1.5 0.5 2.1 1.41.5 2.1 1.4 1.3 1.60.4 0 0 1.0 1.30.6 0.3 0 0.5 0.30.3 0.3 0.5 0 0.31.4 2.4 1.4 0.5 1.4

0.5 00.3 0.30.7 02.5 3.6

1.1 0.32.4 7.40.3 00.8 1.5

2.6 5.32.2 2.12.3 2.1

3.2 0.91.0 0.91.5 0.6

0.7 1.21.2 1.2

1.0 03.2 1.8

1.0 1.22.8 2.7

4.1 3.01.3 2.4

2.5 4.53.0 3.6

4.9 3.91.8 0

0.9 1.0 0.60.4 0.3 0.30.4 0.8 0.42.9 2.1 2.9

0.72.90

1.1

7.73.03.2

1.42.71.3

0.91.3

0.22.1

1.14.5

3.82.7

2.02.5

1.02.90

1.3

8.93.12.1

0.51.60.8

1.01.3

0.31.6

1.83.7

3.72.9

3.92.3

6.1 4.20.4 0.3

TABLE 3-Continued

Amino a

E. coli Z. mobilisbamid Codona combinedbacid (mol%) gap pdc adhB Combinedd

Cys TGT 0.4 0 0.2 0.3 0.2TGC 0.5 0.9 1.1 1.3 1.1

Trp TGG 0.7 0.6 1.1 0 0.6

Arg CGT 3.1 4.2 1.3 1.6 2.4CGC 2.0 1.5 1.6 0.3 1.1CGA 0.2 0 0 0 0CGG 0.2 0 0.2 0 0.1AGA 0.06 0 0 0 0AGG 0.04 0 0 0 0

Gly GGT 3.8 4.2 5.7 6.0 5.3GGC 3.1 3.0 2.0 2.3 2.4GGA 0.4 0.3 0.2 0 0.2GGG 0.6 0 0 0 0

a Preferred codons are underlined.bExpressed as a molar percentage of codon usage.c Combination of 52 proteins in E. coli (2).d Combination of three glycolytic enzymes from Z. mobilis.

gene from Z. mobilis in E. coli. Table 2 summarizes thespecific activity of glyceraldehyde-3-phosphate dehydro-genase in strains of E. coli with and without plasmidscontaining the gap gene from Z. mobilis. Strain DF221exhibited very low levels of glyceraldehyde-3-phosphatedehydrogenase activity, as shown previously (16). Activitywas readily detected in strain TC4, which carries the wild-type gene. The level of glyceraldehyde-3-phosphate dehydro-genase activity in strain DF221 containing pLOI310 (theoriginal 5.5-kb fragment from Z. mobilis) was equivalent tothat of wild-type E. coli, consistent with the observedcomplementation of the gap gene mutation. Tenfold-higheractivity was observed in strain DF221 containing pLOI314,in which the lac promoter of pUC18 was oriented in thedirection of gap transcription. Only 1/20 of this activity wasobserved in a similar subclone (pLOI312) which was ori-ented in the opposite direction with respect to the pUC18 lacpromoter. These results provide evidence that the nativepromoter for gap from Z. mobilis is considerably lessefficient in E. coli than the lac promoter. The specific activityof glyceraldehyde-3-phosphate dehydrogenase in Z. mobilisranges from 0.8 to 4 IU/mg of total cell protein (27). E. coliDF221 containing pLOI314 contained even higher activities.

Figure 2 shows a sodium dodecyl sulfate-polyacrylamidegel of whole cell proteins from Z. mobilis and E. coli. Pawluket al. (29) have reported that the subunit molecular weight ofZ. mobilis glyceraldehyde-3-phosphate dehydrogenase was

36,000, based on gel filtration of native enzyme, although thedenatured subunits migrated in the region corresponding to41,000 on sodium dodecyl sulfate-polyacrylamide gels. Aprominent band in the region corresponding to a molecularweight of 41,000 was clearly visible in Fig. 2 (lane c, strainTC4 containing pLOI314). This strain contained the highestactivity of Z. mobilis glyceraldehyde-3-phosphate dehydro-genase. Much less prominent bands were also present in thisregion in strain DF221 lacking genes from Z. mobilis (lane a)and in E. coli TC4 containing pUC18 (lane b). This band was

less prominent in strain TC4 containing pLOI310 (lane d) and

FIG. 3. Sequence and translation of the glyceraldehyde-3-phosphate dehydrogenase gene from Z. mobilis. The proposed Shine-Dalgarno(S.D.) sequence is overlined and labeled. Two promoters were found. The promoter nearest the Shine-Dalgarno sequence is labeled P2.

Regions corresponding to the -10 and -35 sites are underlined. The distal promoter is labeled P1. Transcriptional initiations by Z. mobilisare indicated by arrows. Transcriptional initiations by E. coli are indicated by the symbol >.

5657VOL. 169, 1987

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

5658 CONWAY ET AL.

P 1-

P2- ....

U._~

,...

_ _

_

a bcdFIG. 4. Primer extension analysis of transcriptional initiation.

Lanes: a, RNA from Z. mobilis CP4; b, RNA from E. coli TC4containing pLOI312; c, RNA from E. coli TC4 containing pLOI314;d, RNA from strain TC4 lacking plasmid. Lanes at the left labeled A,G, T, and C represent a sequencing ladder of the coding strand. Thepositions of the two promoters recognized by Z. mobilis are markedas P1 and P2.

in strain TC4 containing pLOI312 (lane e) than in strainTC4(pLOI314). This putative glyceraldehyde-3-phosphatedehydrogenase band in E. coli corresponded to one of themajor protein bands in Z. mobilis (lane f). From the reportedspecific activity for purified glyceraldehyde-3-phosphatedehydrogenase, 205 IU/mg (29), this enzyme can be calcu-lated to represent 0.1 to 0.2% of the total cell protein in Z.mobilis CP4 and E. coli DF221 containing either pLOI310 or

pLOI312. Similarly, strain DF221 containing pLOI314 can

be estimated to contain over 2% of its total cell protein as

this enzyme. Comparison of these calculated values with thegel profiles suggests that the calculated values may underes-timate the levels of this Z. mobilis protein in recombinantstrains of E. coli.Sequence of the glyceraldehyde-3-phosphate dehydrogenase

gene. Figure 3 shows a summary of the nucleotide andtranslated amino acid sequence for the gap gene from Z.mobilis. This gene contained an open reading frame encod-ing 337 amino acids (including the N-terminal methionine)with an aggregate molecular weight of 36,099. Table 3 showsa summary of codon usage for gap. Seventeen amino acidcodons were unused. A proposed ribosome-binding site(AGGAG) occurred 7 bases upstream from the first in-frameATG. The short intervening region between the ribosome-binding site and the ATG and the 42 bp immediately up-

stream from the proposed ribosome-binding site were par-ticularly rich in adenine and thymidine. The open readingframe was followed by tandem, in-frame termination codonswhich were separated by 18 bp. Two inverted repeats whichwere separated by 3 bp (TTTTTGCGAT/CTT/ATCGGTAAAA) occurred immediately following the first transla-tional terminator. This sequence was followed by a secondpotential hairpin (GGG/TTGT/CCC). Several sequenceswere present within the region 200 bp upstream from theproposed ribosome-binding site which had some homologywith the proposed enteric bacterial consensus sequence (30)for the -10 and -35 regions. The G+C content of the 1,713base pairs sequenced was 50.6%, higher than the reported48.5% G+C composition of bulk genomic DNA from Z.mobilis (24). That of the coding region alone was 53.9%.

Transcriptional initiation of gap in E. coli and Z. mobilis.The sites of transcriptional initiation were identified byprimer extension with mRNA as a template (Fig. 4). At leasttwo and possibly three transcriptional starts appeared to bepresent in mRNA from Z. mobilis. The two most likely startsoccurred over 150 bp upstream from the proposed ribosome-binding site, with one initiation at a thymidine and the otherat an adenosine. The proximal promoter (labeled P2) hasbeen assigned as bp + 1, with the upstream promoter as basepair -27 (labeled P1). The -35 region of the proximalpromoter overlapped the -10 region of the distal promoter(labeled P2). A third band was observed (bp + 121) whichmay represent a third transcriptional initiation or a partialdegradation product.mRNA isolated from E. coli TC4(pLOI312) (Fig. 4, lane b)

appeared to have 5' termini mapping near the regions iden-tified as the two main promoter sites for Z. mobilis. In bothcases, adjacent multiple initiations were observed immedi-ately upstream from the regions identified in Z. mobilis.Such multiple initiations may reflect imprecise binding andinitiation by E. coli RNA polymerase compared with that ofZ. mobilis. No initiation sites were observed in the region ofthe Z. mobilis promoter by using mRNA from strainTC4(pLOI314) (Fig. 4, lane c), the construction with thehighest glyceraldehyde-3-phosphate dehydrogenase activity.In this strain, the only band present was at the top of the gel(not shown) and was presumed to be derived from the lacpromoter. This very active upstream promoter may have

TABLE 4. Amino acid homology amongglyceraldehyde-3-phosphate dehydrogenases from

Z. mobilis and from other organisms

% Identityb

Sequencea Cofactor Catalytic S-loopregion region region Combined

(aal-152) (aa153-336) (aa182-204)

Z. mobilis 100 100 100 100E. coli 48.7 47.3 39.1 47.9B. stearothermophilus 51.3 58.7 60.9 55.4T. aquaticus 46.0 52.2 60.9 49.4S. cerevisiae (tdhl) 49.3 51.6 39.1 50.6Homo sapiens 44.1 44.6 26.1 44.3

(muscle)Nicotiana tabacum

chloroplastgapA 48.0 52.2 60.9gapB 47.4 55.4 60.9a References for sequences are found in the legend to Fig. 5, except for N.

tabacum (33).baa, Amino acid residues.

J. BACTERIOL.

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

Z. MOBILIS GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE

blocked the weaker initiations associated with the Z. mobilispromoters. Similar promoter occlusion has been reported forother systems (4, 31).Comparison of codon usage between the three glycolytic

genes of Z. mobilis and the aggregate codon usage reported forE. coli. Only 39 of the possible 61 amino acid codons wereused more than once by Z. mobilis in the gap gene. Eightcodons were not used in gap or in the two other Z. mobilisgenes (CTA, ATA, GTA, ACA, CGA, AGA, AGG, andGGG). All of these codons except GTA are minor codons inE. coli (12). A single codon was clearly dominant (underlinedin Table 3) for each amino acid except serine, histidine,lysine, and aspartic acid in the Z. mobilis genes. None of thedominant codons in gap, adhB, or pdc represent minorcodons in E. coli. Both E. coli and'Z. mobilis utilized thesame codons most frequently for each amino acid except intwo cases, alanine and tyrosine. In general, the extent ofcodon bias in gap and the other two genes from Z. mobiliswas more severe than for the combined genes of E. coli (2).Such codon preference may represent bias for abundanttRNAs in Z. mobilis, facilitating the biosynthesis (12, 14) ofhigh levels of these three glycolytic enzymes.

Conservation of primary amino acid sequence amongglyceraldehyde-3-phosphate dehydrogenases. The primaryamino acid sequences of glyceraldehyde-3-phosphate de-hydrogenase from all organisms examined show consider-able homology (3, 6, 15), as does the sequence from the Z.mobilis gene. Figure 5 shows a comparison of the primarysequences between Z. mobilis and selected other organisms:E. coli, Bacillus stearothermophilus, Thermus aquaticus, 5.cerevisiae (tdhl), and human muscle. All of these containedlarge regions of functional amino acid homology (indicatedby boxing) and amino acid identity (indicated by shading).Table 4 summarizes the extent of amino acid identity be-tween the Z. mobilis enzyme and the five other sequences.Overall, the Z. mobilis primary sequence showed the highesthomology with that of the B. stearothermophilus enzyme(55.4%). Amino acid positions (Z. mobilis numbers) whichhave been identified as being important in binding andcatalysis (15), such as cysteine-153, glycine-8, aspartate-35,serine-152, threonine-154, and histidine-180, were 'all con-served in Z. mobilis. The lysine-187 found in most eucary-otic sequences and in E. coli (3) has been converted to anarginine in the Z. mobilis sequence, analogous to B. stearo-thermophilus (36) and T. aquaticus (17).The peptide sequence of glyceraldehyde-3-phosphate de-

hydrogenase can be roughly divided (15) into a nicotinamike-binding region, catalytic region, and S-loop region (withinthe catalytic region). This S-loop region is particularly im-portant in that it contains the hydrophobic sequences whichform the core of the tetrameric enzyme and important ioniccontact points between subunits (26). All genes examinedshowed a similar level of homology with Z. mobilis in thecatalytic and nicotinamide-binding regions, except for thecatalytic region of B. stearothermophilus, in which a some-what higher level of homology (58.7%) was observed. TheS-loop regions of E. coli, S. cerevisiae, and human muscleshowed less homology with this region of the Z. mobilisgene. In contrast, the S-loop region of the T. aquaticus andB. stearothermophilus enzymes exhibited 60.9% homologywith that of Z. mobilis. Based on this comparison of primarystructure, the Z. mobilis gene for glyceraldehyde-3-phosphate dehydrogenase appears to be more closely relatedto that of the two thermophilic bacteria than to either that ofE. coli or the two eucaryotic sequences examined. It isinteresting that the chloroplastic glyceraldehyde-3-phos-

phate dehydrogenase isozymes also had more homologywith the enzymes from the two thermophilic bacteria and thechloroplast enzymes than with eucaryotic enzymes (cyto-plasmic) or those of mesophilic bacteria.The S-loop region has been proposed as being particularly

important for thermal stability (26). Despite the apparentsimilarity of peptide sequence between Z. mobilis and thethermophilic bacteria in this region, crude preparations ofthis enzyme from Z. mobilis and from recombinants of E.coli DF221(pLOI314) appeared to be more sensitive tothermal inactivation than the native E. coli enzyme. After 5min of incubation at 60°C followed by immediate cooling to0°C, 51% of the original activity was retained by the nativeE. coli enzyme from strain TC4. The native Z. mobilisenzyme from strain CP4 retained only 20% of the originalactivity after this treatment, and 8% of the original activitywas retained by E. coli DF221 containing pLOI314.Comparison of transcriptional and translational control

sequences between the three glycolytic genes from Z. mobilisand other bacterial sequences. The three glycolytic genesfrom Z. mobilis had several common features. Two of thesehad tandem promoters, gap and adhB. All three containedlong, untranslated 5' leader sequences. Although the func-tion of these leader sequences is unknown, such sequencescould protect mRNA-coding regions to some extent againstdegradation.The Shine-Dalgarno sequences in the glycolytic genes of

Z. mobilis were analogous to those proposed as beingoptimal for E. coli (34). These were AGGAG in gap, GGAGin pdc, and GAGGT in adhB. The spacing between the ATGand these ribosome-binding sequences was 6 to 7 bases,similar to the average for E. coli of 7 + l'base (34). In allthree genes, intervening sequences between the Shine-Dalgarno and ATG were A+T rich, a composition which hasbeen reported to enhance translation rates in E. coli (34). Analanine c'odon (GCT) immediately following the first in-frameATG has been reported to be present in highly expressedgenes in E. ccli (34). An alanine codon (GCG or GCT) waspresent in this position in gap and addhB. Thus, the transla-tional control sequences on the Z. mobilis glycolytic en-zymes contain many of the features of highly expressedgenes in E. coli.The sequences present in the promoter regions of the three

glycolytic genes from Z. mobilis also had some degree ofhomology (Fig. 6A). Although the sample size was quitesmall (five promoters), several features may prove importantfor recognition. The -10 regions of the gap P2 and adhB P2promoters were remarkably similar in sequence, with iden-tity between 7 of 9 consecutive bases. The -10 sequence ofgap P1 had more homology with that of the pdc promoterthan with the gap P2 promoter, with identity between 5 of 9consecutive bases. The -10 region of the adhB P1 promoterwas most similar to that of the gap P1 promoter, withidentity between 5 of 9 consecutive bases. A tentativeconsensus can be deduced in which a position is defined byidentity within 4 of the 5 promoters, 5'-TA**G***T-3'. Theposition of the conserved 5' T ranged from -10 for adhB P2to -13 for pdc with respect to the initiation of transcription(bp + 1). Since our assignment of transcriptional initiation byprimer extension is generally accurate to within 2 bases, thisdiscrepancy in spacing may well be within experimentalerror. The average spacing between the -10 region and thetranslational initiation site was 10.6 bp for the three glyco-lytic enzymes of Z. mobilis. The sequence in the -10 regionresembled the enteric bacterial consensus sequence (31) inthat it began with the highly conserved TA and ended with

VOL. 169, 1987 5659

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

5660 CONWAY ET AL.

12

3

4

5

6

10^V K V A I N G F G R I G RL A

TI K V G I N G F G R I C R I V

J K Y G I N G F G R I G R N V

LVKVGIU-GFGRIGRHVV R Y A IV D F4I R I G R L V

G K V -K G VDI .QLL Vi

1 AJT Y P

2 T ECjR F D

3 VjGR L D

4 I V H R F P

5 T[tR Y A

6 TILK F H

|A E V

C. E V

60T T

.E Vv v

AY

SNKA

E G N D

K DC H

N D G D

D D Q Y

D D K H

OD G K

20 30 40 50

A I S R P D S G L !.-1 L T TN S V E N A F £; R

F R A AQK R S DI1V2I V AIR N D LIL D A D Y MA Y M L YP

F R A A K N P D I A V V A V W. D L|T-N A D G- L| H LLYS

FIR I L RS G - V'. V A L I T N D K T L 4HI. L £Y*

MITI ALS RP NVIVj9V A L FIT N D YA YY Y

T AF N S G K|V D I A P F I D L HYM Y 2myY

70[M Y.:j G K..Mw. SLI V....- . . .R; T:.......L Y V ; A< RAZ.I T-

I IV DKZIAT... ........... t..

L A L T~~~~~~I FC

110 1201 X TKSI -x. .E- - -^. s Y L T :tD2 fr T D EjTjAJRKIII^£^TAGAKZVV MTCG?'KD N3 t KREzDA K N K KL ES-4P1A1LVE.-A.-: I4 T DA D KA^KA E[C*AEKVIZTA:V T NE3 K R . .. . .. . . ......D :

5 PL iD IQIE L D IiDFA KKYVITA?S S T A

6 T MLL AI A KI I S A D A

K KUD EfVG E I

A QA

G S C

G DE

130V D R T

T P M F

N I

DIIP M F

P M F

160-~~~~~~~~~.........

I L L Ki. 71 * --U-Wt-TIr LHTrQQ-KIF. sv

2 I&+*r%tt*CLAP--L A K I N D N F I E3 frSCWTBCL}.-i A K tHQE FIGIJI4 ~I#1?TISLAPVMJ E E AF VJE

65ILL IIISj.2JLL A K DIHDHFEuE6~~~~~~~~~~~A H D H F E

190FQI jS#

T A T K T V GPS KmY :i:- N1i^I1LP-U K#

L LP-j Kz

L A T-Q K T V GPS K

I A T TV SPS G K L

1 TS ASA2C::A.f

3 A A ES I:.

AA3 A I N I$

4 AAA G NjII6 A A Q NLJ

230|gI SWe

SeIT N A

T M A

TM AIT ;M A

[T M A

L FF V P

:L R L

V L V A E LVL ALR L

fLI V K L

L C R L

260RFDT T: SJ o- K A-A- D

K A T YIIIIIL A A V KA AA EK ElvT v it9 v VIA AL K A A A EIrElVTAI IVAAAALKAAA E

EKT

T YID :.IK K V V : A A A EIPA K YLD D JK K V E S

T270

|E L

|a LiP U

T

K

K

K

K

K

IT;l; IX E D

I A I?: S E IA

EI :LAA

:....

A

E

Y s

|V.$|TEMN G E VC|L?SR::w FN |Y N S T V

SS Cv

D NCG

la0TF EECKI L AR

Cl I A Li N D N F V lK LY|S

Tjf,L|CN M- -|KF,AIG ItQiLiS P K F V|K L VSGTIE NDTFV SI

TD)N E

VYTD NIE

W D N

::eYDN I:

Delw

T

T

Fw.y

Gy..

|G Y S H|

1 Y.P

C Y ST

E

'.SM SVK V L D

..i

V A D

v v D

A-Y Dp

T

L

L

L

L

L

336A QAX K T (END)

I A HI SjK (END)

A A YIN A:Al L (END)

V Q LVL Ri GV (END)

V E H I AK A (END)

H A S K E (END)

250QE

EK

D

E

2

3

4

5

6

S R

V F A K

FI V;t A L

A K

I F A S

I F WA G

300E

A

SL

A

A

2

3

4

5

6

*7w1 .. M-MI. r.9aA_ ,IIAl'II1.1

J. BACTERIOL.

2

,_K

3

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

Z. MOBILIS GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE

A- 10 Region mRNA init. Gene/Promoter

---------------------------------------------------------__-

G T A T T G A T G T (2) T T G >>> 112 P2

A T A T eG C T T T (0) T G C >>> adhB P2

A T A C T G G A A T (2) A T G >> gap P1

T T A G A G G T T T (3) G G T >>> pdc P

G T A G A A A A A T (1) T C G >>> adhB Pi

T A T T G * T y T 2a P2 + adhB P2

T A **G G **T gap Pl + p P

T A* * * A A T gap P1 + adnB Pl

T A * * G * * * T Possible ZM consensus

T A t a a T Enter ic consensus

B-35 Region Intervening Gene/Promoter

---A A G C A G G G A C G A C --- 21 bp gap Pi

---A T A C T G G A A T A A A --- 18 bp 9ap P2

---A G C C T T G C T C A T C 17 bp adhB Pl

---C C C C T T G A T C T G A 18 bp adhB P2

---T G C C T A T A G C T A A 19 bp pdc P

p C * G G p A y p A P + P2

C C T T G * T C adhB Pl + P2

C C T * * * * C

* C 1' * G * * C

adhB + pdc

Possible Zm consensus

r T G a c a t (17 bp) Enteric consensus

FIG. 6. Comparison of sequences in promoter regions among thethree glycolytic enzymes of Z. mobilis (Zm). (A) -10 region; (B)-35 region. Numbers in parentheses in panel A refer to additionalbases which separate the sequence in the -10 region from the baseposition at which mRNA synthesis appears to be initiated (under-lined). Dashes denote additional flanking sequences which areuntranscribed, and >>> represents additional sequences which arepart of the mRNA. A, T, G, and C, base sequence or most highlyconserved bases in promoter comparisons; a, t, g, and c, baseswhich are usually conserved; *, variable bases; y, pyrimidine; p,purine.

the highly conserved T, although the Z. mobilis sequenceappeared to be longer.We have tentatively identified a -35 region in the pro-

moter regions of the Z. mobilis glycolytic genes (Fig. 6B).The -35 regions of the two gap promoters were homologousin 5 of 9 consecutive bases. The -35 regions of the two adhBpromoters were hormologous in 7 of 8 consecutive bases. The-35 region of pdc most closely resembled that of the adhBgene, sharing 4 of 8 consecutive bases. A tentative consen-sus sequence can be deduced in which a position is assignedby identity in 4 of the 5 promoters, *CT*G**C. The numberof intervening bases between the 3' conserved C of the -35region and the 5' conserved T of the -10 region ranged from

17 to 21 bp, with an average of 18.6 bp. This sequence in the-35 region had some homology with the consensus se-quence proposed for E. coli (31). The differences betweenthe Z. mobilis promoter sequences and the consensus se-quence for E. coli (31) in the -10 and -35 regions of gap areconsistent with our observations that these native sequenceswere less efficient than the enteric bacterial lac promoter onderivatives of pUC18. It is interesting that the -35 regionswithin the gap and adhB genes which contained tandempromoters were more similar to each other than to those ofother genes. In contrast, the -10 regions were more similarbetween the two distal promoters and between the twoproximal promoters of adhB and gap than they were withina single gene.

ACKNOWLEDGMENTS

We appreciate the help of B. Bachmann in providing mutantstrains of E. coli and advice concerning their handling. We alsothank K. T. Shanmugam and W. E. Gurley for their suggestionsduring the preparation of the manuscript.These studies were supported in part by grants from the National

Science Foundation (DMB 8204928), by the Department of Energy,Office of Basic Energy Sciences (DE-FG05-86ER13575), by the U.S.Department of Agriculture, Alcohol Fuels Program (86-CRCR-1-2134), and by the Florida Agricultural Experiment Station (no.8179).

LITERATURE CITED

1. Algar, E. M., and R. K. Scopes. 1985. Studies on cell-freemetabolism: ethanol production by extracts of Zymomonasmobilis. J. Biotechnol. 2:275-287.

2. Allf-Steinberger, C. 1984. Evidence for coding pattern on thenon-coding strand of the E. coli genome. Nucleic Acids Res.12:2235-2241.

3. Branlant, G., and C. Branlant. 1985. Nucleotide sequence of theEscherichia coli gap gene. Different evolutionary behavior ofthe NAD+-binding domain and the catalytic domain of D-glyceraldehyde-3-phosphate dehydrogenase. Eur. J. Biochem.150:61-66.

4. Busby, S., H. Aiba, and B. DeCrombrugghe. 1982. Mutations inthe Escherichia coli operon that define two promoters and thebinding site of the cyclic AMP receptor protein. J. Mol. Biol.154:211-227.

5. Byers, L. D. 1982. Glyceraldehyde-3-phosphate dehydrogenasefrom yeast. Methods Enzymol. 89:326-335.

6. Chen, S., T. D. Lee, K. Legesse, and J. E. Shively. 1986.Identification of the arylazido-p-alanyl-NADW-modified site inrabbit muscle glyceraldehyde-3-phosphate dehydrogenase bymicrosequencing and fast atom bombardment mass spectrome-try. Biochemistry 25:5392-5395.

7. Conway, T., M. O.-K. Byun, and L. 0. Ingram. 1987. Expres-sion vector for Zymomonas mobilis. Appl. Environ. Microbiol.53:235-241.

8. Conway, T., Y. A. Osman, and L. 0. Ingram. 1987. Geneexpression in Zymomonas mobilis: promoter struct,ire andidentification of membrane anchor sequences forming functionallacZ' fusion proteins. J. Bacteriol. 169:2327-2335.

9. Conway, T., Y. A. Osman, J. I. Konnan, E. M. Hoffmann, andL. 0. Ingram. 1987. Promoter and nucleotide sequences of theZymomonas mobilis pyruvate decarboxylase. J. Bacteriol. 169:949-954.

10. Conway, T., G. W. Sewell, Y. A. Osman, and L. 0. Ingram.1987. Cloning and sequencing of the alcohol dehydrogenase IIgene from Zymomonas mobilis. J. Bacteriol. 169:2591-2597.

FIG. 5. Comparison of amino acid homology between selected glyceraldehyde-3-phosphate dehydrogenases from different organisms.Major regions of conserved functional homology are boxed, and amino acids with identity are shaded. Lines: 1, Z. mobilis; 2, E. coli (3); 3,B. stearothermophilus (36); 4, Thermus aquaticus (17); 5, S. cerevisiae (tdhl) (18); 5, Homo sapiens muscle (25).

VOL. 169, 1987 5661

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

5662 CONWAY ET AL.

11. Dawes, E. A., and D. W. Ribbons. 1966. Sucrose utilization byZymomonas mobilis: formation of a levan. Biochem. J. 98:804-812.

12. de Boer, H. A., and R. A. Kastelein. 1986. Biased codon usage:an exploration of its role in optimization of translation, p.225-285. In W. Reznikoff and L. Gold (ed.), Maximizing geneexpression. Butterworth Publishers, Stoneham, Mass.

13. Doimbek, K. M., and L. 0. Ingram. 1986. Magnesium limitationand its role in the apparent toxicity of ethanol during yeastfernmentation. Appl. Environ. Microbiol. 52:975-981.

14. Grosjean, H., and W. Fiers. 1982. Preferential codon usage inprokaryotic genes: the optimal codon-anticodon interactionenergy and the selective codon usage in efficiently expressedgenes. Genes 18:199-209.

15. Harris, J. I., and M. Waters. 1976. Glyceraldehyde-3-phosphatedehydrogenase, p. 1-49. In P. D. Boyer (ed.), The enzymes,vol. XIII. Academic Press, Inc., New York.

16. Hillman, J. D., and D. G. Fraenkel. 1975. Glyceraldehyde-3-phosphate dehydrogenase mutants of Escherichia coli. J. Bac-teriol. 122:1175-1179.

17. Hocking, J. D., and J. I. Harris. 1980. D-Glyceraldehyde-3-phosphate dehydrogenase. Amino acid sequence of the enzymefrom the extreme thermophile, Thermus aquaticus. Eur. J.Biochem. 108:567-579.

18. Holland, M. J., and J. P. Holland. 1979. The primary structureof a glyceraldehyde-3-phosphate dehydrogenase gene from Sac-charomyces cerevisiae. J. Biol. Chem. 254:9839-9845.

19. Layne, E. 1957. Spectrophotometric and turbidometric methodsfor measuring proteins. Methods Enzymol. 3:447-454.

20. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193:265-275.

21. Luria, S. E., and M. Delbruck; 1943. Mutations of bacteria fromvirus sensitivity to virus resistance. Genetics 28:491-511.

22. Mandel, M., and A. Higa. 1970. Calcium dependent bacterio-phage DNA infection. J. Mol. Biol. 53:159-162.

23. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

24. Montenecourt, B. S. 1985. Zymomonas, a uniqUe genus ofbacteria, p. 261-289. In A. L. Demain and N. A. Solomon (ed.),

Biology of industrial microorganisms. Benjamin-CummingsPublishing Co., Menlow Park, Calif.

25. Nowak, K., M. Wolny, and T. Banas. 1981. The complete aminoacid sequence of human muscle glyceraldehyde-3-phosphatedehydrogenase. FEBS Lett. 134:143-146.

26. Olsen, K. W. 1983. Structural basis for the thermal stability ofglyceraldehyde-3-phosphate dehydrogenases. Int. J. PeptideProtein Res. 22:469-475.

27. Osman, Y. A., T. Conwvay, S. J. Imonnetti, and L. 0. Ingram.1987. Glycolytic flux in Zymomonas mobilis: enzyme and me-tabolite levels during batch fermentation. J. Bacteriol. 169:3726-3736.

28. Osman, Y. A., and L. 0. Ingram. 1985. Mechanism of ethanolinhibition of fermentation in Zymomonas mobilis strain CP4. J.Bacteriol. 164:173-180.

29. Pawluk, A., R. K. Scopes, and K. Griffiths-Smith. 1986. Isolationand properties of the glycolytic enzymes from Zymomonasmobilis. Biochem. J. 238:275-281.

30. Pustell, J., and F. C. Kafatos. 1984. A convenient and adaptablepackage of computer programs for DNA and protein sequencemanagement, analysis, and homology determination. NucleicAcids Res. 12:643-655.

31. Reznikoff, W. S., and W. R. McClure. 1986. E. coli promoters,p. 1-33. In W. Reznikoff and L. Gold (ed.), Maximizing geneexpression. Butterworth Publishers, Stoneham, Mass.

32. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

33. Shih, M.-C., G. Lazar, and H. M. Goodman. 1986. Evidence infavor of the symbiotic origin of chloroplasts: primary structureand evolution of tabacco glyceraldehyde-3-phosphate dehydro-genases. Cell 47:73-80.

34. Stormo, G. D. 1986. Translation intiation, p. 195-224. In W.Reznikoff and L. Gold (ed.), Maximizing gene expression.Butterworth Publishers, Stoneham, Mass.

35. Swings, J., and J. DeLey. 1977. The biology of Zymomonas.Bacteriol. Rev. 41:1-46.

36. Walker, J. E., A. F. Carne, M. J. Runswick, J. Bridgen, and J. I.Harris. 1980. Glyceraldehyde-3-phosphate dehydrogenase. Com-plete amino-acid sequence of the enzyme from Bacillus stearo-thermophilus. Eur. J. Biochem. 108:549-565.

J. BACTERIOL.

on February 14, 2018 by guest

http://jb.asm.org/

Dow

nloaded from